Koofers

Exam 3 - Flashcards

Flashcard Deck Information

Class:MKT 431 - MARKETING RESEARCH
Subject:MARKETING
University:Clemson University
Term:Spring 2012
- of -
INCORRECT CORRECT
- INCORRECT     - CORRECT     - SKIPPED
Shuffle Remaining Cards Show Definitions First Take Quiz (NEW)
Hide Keyboard shortcuts
Next card
Previous card
Mark correct
Mark incorrect
Flip card
Start Over
Shuffle
      Mode:   CARDS LIST       ? pages   PRINT EXIT
primary data data you collect yourself
secondary data data that already exists
applications of secondary data economic forecast
corporate intelligence
public opinion
historical data

internal secondary data exists within a compnay
Generated by Koofers.com
external secondary data exists outside a company
database marketing making marketing decisions based on company databases
creating customer databases to build relationships
CRM customer relationship management
data mining digging through information to find important data
  • use regression
  • looking for relationhips
Generated by Koofers.com
databases collection of data 
records rows
fields columns
sources sales records and reports
scanner data 
Generated by Koofers.com
external secondary data published
syndicated services
external databases
published publicly distributed (libraries, trade organizations
syndicated services data collected in standard format
subscription based
specialized
ex arbitron and nielsen
advantages of secondary data quick, readily available
inexpensive
enhance primary data
may achieve objectvie
Generated by Koofers.com
disadvantages of secondary data incompatible measurement units and class definitions
  • need zip code data given county data
outdated
how to evaluate secondary data purpose of study
who collected the data
whats being measured
how is it being measured
consistent with other information
boolean logic and, or, not
field searching author, title, subject
Generated by Koofers.com
proximity operators adjacent, near, same
truncation search on a root word
nesting search order
limiting meet specified criteria
Generated by Koofers.com
secondary data sources population census
statistical abstract of the US
NAICS North American Industry Classification System
Business coding
general and specific
20 broad sectors identified by 6 digit codes
survey of buying power published in "Sales and Marketing Management"
buying power index
Buying power index % of US buying power in a market
BPI= ((population of Market area/ total us pop) x2) + ((effiective buying income of market area/total us ebi) x5) +( (retail sales of market area/ total us retail sales) x3)
Generated by Koofers.com
effective buying income disposable income
gross income- taxes
lifestyle market analyst market profiles
lifestyle profiles
DMA designated market area
total population, jobs, education, race, age, income
standardizes information data collected, or data collection process, standardized for all users
Generated by Koofers.com
syndicated data collect data in standard format, available to subscribers
Nieson tv ratings
standardized services standardized process generates information
ESRI's community tapestry 
advantages of standardized services shared costs
high quality
quickly disseminate
disadvantages of standardized services buyers have little control
long term contracts 
no strategic advantage
Generated by Koofers.com
consumer attitudes and opinion polls Yankelovich monitor, harris poll, gallup poll
geodemographics classification of geographic areas based on characteristics of inhabitants
PRIZM neilsen
66 clusters
Community tapestry ESRI
65 clusters
Generated by Koofers.com
market tracking monitor variable over time
sales trhgouh food, drug, mass merchandise stores
Nielsen: scan track (weekly)
IRI: infoscan custom store tracking
in store audits
rating % of all household watching show
share % of all households with tv turned on watching show
people meters 25000 metered households
Generated by Koofers.com
diaries 16 million diaries during sweeps
arbitron radio
MRI's starch readership service noted
associated
read some
read most
simmons national consumer study multimedia
Generated by Koofers.com
single source data info on multiple variables
recorded from respondent panel
test causal relationships
qualitative less structured, info from less people
quantitative questionnaire, more structured info from more people
pluralistic research do both together
qualitative before quantitative
Generated by Koofers.com
direct observation behavior in real time
indirect observation results of behavior
  • archives
  • physical traces
disguised observation subject unaware
eg mystery shopper
undisguised observation subject aware
eg manager evaluating sales techniques
Generated by Koofers.com
structured observation researchers uses a checklist
unstuctured observation all behavior monitored
human observation person observes behavior
mechanical observation static observing device
Generated by Koofers.com
when is observation appropriate short duration
public
faulty, deceitful recall
advantages of observation insight into actual behaviors
no recall error, accurate
low cost (sometimes)
disadvantages to observation small number of subjects
subjective interpretation
don't know "why"
focus groups small discussion group guided by moderator
Generated by Koofers.com
moderator encourages discussion and makes sure its focused
generate ideas, learn vocabulary, insights into needs and attitudes
traditional focus groups 6 to 12 people, dedicated room, one way mirror (possibly) 1-2 hours
non traditional focus group online
  • no physcial setup
  • transcript captured
  • particpants widely separated
  • comfortable environement
  • moderator can chat with individuals
  • may lose informaton
advantages of focus group generate new ideas
observce particpants
access to special groups
Generated by Koofers.com
disadvantages of focus groups representative (how representative is it really)
diffiucult interpretation
high costs 
depth interview one on one questions to understand attitudes and behavior
laddering attributes -> consequences -> values
protocol analysis respondent verbalizes decision making process
Generated by Koofers.com
types of projective techniques simulated actives to get answers not revealed under direct questioning
  • word association
  • sentence completion
  • picture test
  • carbon/balloon test
  • role-playing 
ethnographic research descriptive study of group
observe, interview, record behavior
researchers shouldn't interfere with behavior
physiological measurements monitoring involuntary responses to stimuli
  • pupilometer
  • galvanometer
  • ct scans

surveys predesigned questionnaire for many respondents
standardized
easy to administer
tap "unseen"
suitable for analysis
identify group differences
Generated by Koofers.com
person administered surveys interviewer reads questions and records answers
face to face or telephone
advantages of person administered surveys feedback
rapport
quality control
adaptable

disadvantages of person administered surveys human error
slow
costly 
interview evaluation

computer assisted surveys computer technology plays some role 

Generated by Koofers.com
advantages of computer assisted surveys speed
fewer errors
graphics and videos
low "interview evaluation" concerns

disadvantages of computer assisted surveys technical skills
set up costs
self administered survey respondent completes questionnaire alone
  • paper and pencil

advantages of self administered surveys cost
respondent sets pace
low evaluation concerns

Generated by Koofers.com
disadvantages of self administered surveys respondent controls survey
no one to clarify/encourage questions
must be perfect
mixed mode (hybrid) surveys multiple data collection methods
advantages of mixed mode surveys reach people with various techniques
disadvantages of mixed mode surveys mode may affect response 
more complex
Generated by Koofers.com
which survey method goal of high quality 
balance data quality against
  • time horizon
  • budget
  • incidence rate
  • cultural infrastructure
  • type of interaction required
population group under study
define specifically based on objectives
sample subset of population
where we get info from


sample unit basic level of investigation
Generated by Koofers.com
census accounting of entire population
sampling error occurs because sample is used
sample frame frame doesn't include entire population
sample frame error because frame doesn't include entire population, there is going to be an error of some form
Generated by Koofers.com
reasons to use a sample cost
time considerations
too much data with census
samples can be precise
probability sample known chance of selection
simple random sampling
  • blind draw
  • random numbers
need
  • complete list of population
  • unique IDs for everyone
nonprobability not based on fairness, equity, equal chance
often not representative
types of probability systematic sampling
cluster sampling
stratified sampling
Generated by Koofers.com
systematic sampling more efficident than SRS
need
  • random starting point
  • skip interval = pop size/ sample size
cluster sampling population divided into representative clusters
one step area sample
  • select one area, conduct census
two step area sample
  • reandolly select areas and people in areas
specification error
  • clusters dont represent population

stratified sampling use if answers research questions (likely to vary)
separate population into strata
  • sample from each stratum
  • classificiation measures create strata
  • proportionate stratified sample
  • disproportionate stratified sample
  • larger samples taken from strata with more varaince

nonprobability samples convenience sample
judgement sample
referral (snowball sample)
quata sample
Generated by Koofers.com
convenience sample error: population members don't frequent location from which sample is drawn
judgement sample educated guess as to who represents population
subjective
referral (snowball) sample respondents name other respondents
less known, disliked
  • low probability of being sleected
quata sample specific number of certain types of people
Generated by Koofers.com
online sampling techniques random intercept
invitation
panel
steps in the sample plan define population
obtain list of population
determine sample size, method
draw sample
validate sample
resample if necessary
sample size axioms census: only perfectly accurate sample
random sample: always has error
larger random samples: more accurate
random sample can be tiny % of population and still be accurate
how to determine sample error sampling error: sampling selection and sample size
related to
  • condifence level
  • population variability
  • desired accuracy (acceptable error)
Generated by Koofers.com
central limit theorem regardless of populations distribution means from large samles have normal distribution
factors related to sample size confidence level
population variability
desired accuracy
other methods of determining sample size rule of thumb
conventional
  • previous study
  • ignores current cicrumstances
statistical analysis
  • analysis requires minimum sample size
cost basis
  • all you can afford
  • determined by budget, ignores value of results
examples of nonsampling errors nonresponse
data gathering handling analysis
interpretation
Generated by Koofers.com
intentional fieldworker errors cheating
false information
leading respondents
unintentional fieldworker errors personal characteristics (gender, accents)
misunderstanding (when administering questionnaire)
fatigue
intentional respondent errors falsehoods
nonresponse
unintentional respondent errors respondent misunderstanding
guessing
attention loss/ fatigue
distractions
Generated by Koofers.com
ways to control fieldwork errors supervison, validation
selection, training, role playing
breaks, reduce monotony
controlling respondent errors anonymith, confidientiality, incentives, validation, third person techniques
questionnaire/question design, response options, propmters, reverse endpoints
online data collection errors nonresponse
  • refusal
  • break offs
  • item omissions
compled interview  must be defined
  • useful to report response rates
response rates % of total sample who completed questionnaire

Generated by Koofers.com
components of CASRO formula completions, ineligible, refusals, no answers
reducing nonresponse advance notificatoin
incentives
follow up
questionnaire screening look for
  • incomplete questionnaires
  • item omissions
  • response pattersn
  • yay or nay saying, middle of road
  • unreliable responses
Generated by Koofers.com

List View: Terms & Definitions

  Hide All 127 Print
 
Front
Back
 primary datadata you collect yourself
 secondary datadata that already exists
 applications of secondary dataeconomic forecast
corporate intelligence
public opinion
historical data

 internal secondary dataexists within a compnay
 external secondary dataexists outside a company
 database marketingmaking marketing decisions based on company databases
creating customer databases to build relationships
 CRMcustomer relationship management
 data miningdigging through information to find important data
  • use regression
  • looking for relationhips
 databasescollection of data 
 recordsrows
 fieldscolumns
 sourcessales records and reports
scanner data 
 external secondary datapublished
syndicated services
external databases
 publishedpublicly distributed (libraries, trade organizations
 syndicated services datacollected in standard format
subscription based
specialized
ex arbitron and nielsen
 advantages of secondary dataquick, readily available
inexpensive
enhance primary data
may achieve objectvie
 disadvantages of secondary dataincompatible measurement units and class definitions
  • need zip code data given county data
outdated
 how to evaluate secondary datapurpose of study
who collected the data
whats being measured
how is it being measured
consistent with other information
 boolean logicand, or, not
 field searchingauthor, title, subject
 proximity operatorsadjacent, near, same
 truncationsearch on a root word
 nestingsearch order
 limitingmeet specified criteria
 secondary data sourcespopulation census
statistical abstract of the US
 NAICSNorth American Industry Classification System
Business coding
general and specific
20 broad sectors identified by 6 digit codes
 survey of buying powerpublished in "Sales and Marketing Management"
buying power index
 Buying power index% of US buying power in a market
BPI= ((population of Market area/ total us pop) x2) + ((effiective buying income of market area/total us ebi) x5) +( (retail sales of market area/ total us retail sales) x3)
 effective buying incomedisposable income
gross income- taxes
 lifestyle market analystmarket profiles
lifestyle profiles
 DMAdesignated market area
total population, jobs, education, race, age, income
 standardizes informationdata collected, or data collection process, standardized for all users
 syndicated datacollect data in standard format, available to subscribers
Nieson tv ratings
 standardized servicesstandardized process generates information
ESRI's community tapestry 
 advantages of standardized servicesshared costs
high quality
quickly disseminate
 disadvantages of standardized servicesbuyers have little control
long term contracts 
no strategic advantage
 consumer attitudes and opinion pollsYankelovich monitor, harris poll, gallup poll
 geodemographicsclassification of geographic areas based on characteristics of inhabitants
 PRIZMneilsen
66 clusters
 Community tapestryESRI
65 clusters
 market trackingmonitor variable over time
sales trhgouh food, drug, mass merchandise stores
Nielsen: scan track (weekly)
IRI: infoscan custom store tracking
in store audits
 rating% of all household watching show
 share% of all households with tv turned on watching show
 people meters25000 metered households
 diaries16 million diaries during sweeps
 arbitronradio
 MRI's starch readership servicenoted
associated
read some
read most
 simmons national consumer studymultimedia
 single source datainfo on multiple variables
recorded from respondent panel
test causal relationships
 qualitativeless structured, info from less people
 quantitativequestionnaire, more structured info from more people
 pluralistic researchdo both together
qualitative before quantitative
 direct observationbehavior in real time
 indirect observationresults of behavior
  • archives
  • physical traces
 disguised observationsubject unaware
eg mystery shopper
 undisguised observationsubject aware
eg manager evaluating sales techniques
 structured observationresearchers uses a checklist
 unstuctured observationall behavior monitored
 human observationperson observes behavior
 mechanical observationstatic observing device
 when is observation appropriateshort duration
public
faulty, deceitful recall
 advantages of observationinsight into actual behaviors
no recall error, accurate
low cost (sometimes)
 disadvantages to observationsmall number of subjects
subjective interpretation
don't know "why"
 focus groupssmall discussion group guided by moderator
 moderatorencourages discussion and makes sure its focused
generate ideas, learn vocabulary, insights into needs and attitudes
 traditional focus groups6 to 12 people, dedicated room, one way mirror (possibly) 1-2 hours
 non traditional focus grouponline
  • no physcial setup
  • transcript captured
  • particpants widely separated
  • comfortable environement
  • moderator can chat with individuals
  • may lose informaton
 advantages of focus groupgenerate new ideas
observce particpants
access to special groups
 disadvantages of focus groupsrepresentative (how representative is it really)
diffiucult interpretation
high costs 
 depth interviewone on one questions to understand attitudes and behavior
 ladderingattributes -> consequences -> values
 protocol analysisrespondent verbalizes decision making process
 types of projective techniquessimulated actives to get answers not revealed under direct questioning
  • word association
  • sentence completion
  • picture test
  • carbon/balloon test
  • role-playing 
 ethnographic researchdescriptive study of group
observe, interview, record behavior
researchers shouldn't interfere with behavior
 physiological measurementsmonitoring involuntary responses to stimuli
  • pupilometer
  • galvanometer
  • ct scans

 surveyspredesigned questionnaire for many respondents
standardized
easy to administer
tap "unseen"
suitable for analysis
identify group differences
 person administered surveysinterviewer reads questions and records answers
face to face or telephone
 advantages of person administered surveysfeedback
rapport
quality control
adaptable

 disadvantages of person administered surveyshuman error
slow
costly 
interview evaluation

 computer assisted surveyscomputer technology plays some role 

 advantages of computer assisted surveysspeed
fewer errors
graphics and videos
low "interview evaluation" concerns

 disadvantages of computer assisted surveystechnical skills
set up costs
 self administered surveyrespondent completes questionnaire alone
  • paper and pencil

 advantages of self administered surveyscost
respondent sets pace
low evaluation concerns

 disadvantages of self administered surveysrespondent controls survey
no one to clarify/encourage questions
must be perfect
 mixed mode (hybrid) surveysmultiple data collection methods
 advantages of mixed mode surveysreach people with various techniques
 disadvantages of mixed mode surveysmode may affect response 
more complex
 which survey methodgoal of high quality 
balance data quality against
  • time horizon
  • budget
  • incidence rate
  • cultural infrastructure
  • type of interaction required
 populationgroup under study
define specifically based on objectives
 samplesubset of population
where we get info from


 sample unitbasic level of investigation
 censusaccounting of entire population
 sampling erroroccurs because sample is used
 sample frameframe doesn't include entire population
 sample frame errorbecause frame doesn't include entire population, there is going to be an error of some form
 reasons to use a samplecost
time considerations
too much data with census
samples can be precise
 probability sampleknown chance of selection
simple random sampling
  • blind draw
  • random numbers
need
  • complete list of population
  • unique IDs for everyone
 nonprobabilitynot based on fairness, equity, equal chance
often not representative
 types of probabilitysystematic sampling
cluster sampling
stratified sampling
 systematic samplingmore efficident than SRS
need
  • random starting point
  • skip interval = pop size/ sample size
 cluster samplingpopulation divided into representative clusters
one step area sample
  • select one area, conduct census
two step area sample
  • reandolly select areas and people in areas
specification error
  • clusters dont represent population

 stratified samplinguse if answers research questions (likely to vary)
separate population into strata
  • sample from each stratum
  • classificiation measures create strata
  • proportionate stratified sample
  • disproportionate stratified sample
  • larger samples taken from strata with more varaince

 nonprobability samplesconvenience sample
judgement sample
referral (snowball sample)
quata sample
 convenience sampleerror: population members don't frequent location from which sample is drawn
 judgement sampleeducated guess as to who represents population
subjective
 referral (snowball) samplerespondents name other respondents
less known, disliked
  • low probability of being sleected
 quata samplespecific number of certain types of people
 online sampling techniquesrandom intercept
invitation
panel
 steps in the sample plandefine population
obtain list of population
determine sample size, method
draw sample
validate sample
resample if necessary
 sample size axiomscensus: only perfectly accurate sample
random sample: always has error
larger random samples: more accurate
random sample can be tiny % of population and still be accurate
 how to determine sample errorsampling error: sampling selection and sample size
related to
  • condifence level
  • population variability
  • desired accuracy (acceptable error)
 central limit theoremregardless of populations distribution means from large samles have normal distribution
 factors related to sample sizeconfidence level
population variability
desired accuracy
 other methods of determining sample sizerule of thumb
conventional
  • previous study
  • ignores current cicrumstances
statistical analysis
  • analysis requires minimum sample size
cost basis
  • all you can afford
  • determined by budget, ignores value of results
 examples of nonsampling errorsnonresponse
data gathering handling analysis
interpretation
 intentional fieldworker errorscheating
false information
leading respondents
 unintentional fieldworker errorspersonal characteristics (gender, accents)
misunderstanding (when administering questionnaire)
fatigue
 intentional respondent errorsfalsehoods
nonresponse
 unintentional respondent errorsrespondent misunderstanding
guessing
attention loss/ fatigue
distractions
 ways to control fieldwork errorssupervison, validation
selection, training, role playing
breaks, reduce monotony
 controlling respondent errorsanonymith, confidientiality, incentives, validation, third person techniques
questionnaire/question design, response options, propmters, reverse endpoints
 online data collection errorsnonresponse
  • refusal
  • break offs
  • item omissions
compled interview  must be defined
  • useful to report response rates
 response rates% of total sample who completed questionnaire

 components of CASRO formulacompletions, ineligible, refusals, no answers
 reducing nonresponseadvance notificatoin
incentives
follow up
 questionnaire screeninglook for
  • incomplete questionnaires
  • item omissions
  • response pattersn
  • yay or nay saying, middle of road
  • unreliable responses
36, "/var/app/current/tmp/"